An assessment of attenuation correction of SPECT MPI images generated by deep learning model

This article evaluates the effectiveness of using a deep learning network model to generate reliable attenuation corrected the single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI). The authors collected myocardial perfusion imaging data of 88 patients from a SPECT/CT machine, with an average age of 62.47 years. Then, two datasets are created from the original data: set A includes the deep learning-based attenuation corrected images (Generated Attenuation Correction - GenAC), and the non-attenuation corrected images; set B contains only non-attenuation corrected images. These datasets were diagnosed by two doctors (in which, one has 7 years of experience and the other has 10 years of experience in reading SPECT MPI). The doctors diagnose based on the image data without knowing which dataset it belongs to. The sensitivity, specificity, diagnostic accuracy, and lesion rate were evaluated between the two data sets. Results: The average specificity, sensitivity, and accuracy of the set with the deep learning-based attenuation corrected images were 0.87, 0.86, 0.86, while the results with the non-attenuation corrected images are 0.69, 0.83, and 0.78.

Spectroscopically Identified Emission Line Galaxy Pairs in the WISP Survey*

We identify a sample of spectroscopically measured emission line galaxy (ELG) Pairs up to z = 1.6 from the Wide Field Camera 3 (WFC3) Infrared Spectroscopic Parallels (WISP) survey. WISP obtained slitless, near-infrared grism spectroscopy along with direct imaging in the J and H bands by observing in the pure-parallel mode with the WFC3 on board the Hubble Space Telescope. From our search of 419 WISP fields covering an area of ∼0.5 deg2, we find 413 ELG pair systems, mostly H α emitters. We then derive reliable star formation rates (SFRs) based on the attenuation-corrected H α fluxes. Compared to isolated galaxies, we find an average SFR enhancement of 40%–65%, which is stronger for major Pairs and Pairs with smaller velocity separations (Δ v < 300 km s−1). Based on the stacked spectra from various subsamples, we study the trends of emission line ratios in pairs, and find a general consistency with enhanced lower ionization lines. We study the pair fraction among ELGs, and find a marginally significant increase with redshift f ∝ (1 + z) α , where the power-law index α = 0.58 ± 0.17 from z ∼ 0.2 to ∼1.6. The fraction of active galactic nuclei is found to be the same in the ELG Pairs as compared to the isolated ELGs.

Comparison of PSF and Non PSF BLOB based reconstructed image in Time-of-Flight PET/CT

Hybrid PET/CT imaging with the use of 18F FDG is a widely used imaging technique with major indications in oncology for staging, re-staging and monitoring response to therapy. There is a major issue of partial volume effect in PET images which affects image quality as well as quantitative accuracy in small lesions. Multiple attempts have been made to resolve these issues. The aim of our study was to look into impact of Point‐spread-function (PSF) on reconstructed attenuation corrected (AC) images of PET/CT and to find out best combination of the number of PSF iterations with regularization level while applying PSF.

Harmonization of PET image reconstruction parameters in simultaneous PET/MRI

Objective Simultaneous PET/MRIs vary in their quantitative PET performance due to inherent differences in the physical systems and differences in the image reconstruction implementation. This variability in quantitative accuracy confounds the ability to meaningfully combine and compare data across scanners. In this work, we define image reconstruction parameters that lead to comparable contrast recovery curves across simultaneous PET/MRI systems. Method The NEMA NU-2 image quality phantom was imaged on one GE Signa and on one Siemens mMR PET/MRI scanner. The phantom was imaged at 9.7:1 contrast with standard spheres (diameter 10, 13, 17, 22, 28, 37 mm) and with custom spheres (diameter: 8.5, 11.5, 15, 25, 32.5, 44 mm) using a standardized methodology. Analysis was performed on a 30 min listmode data acquisition and on 6 realizations of 5 min from the listmode data. Images were reconstructed with the manufacturer provided iterative image reconstruction algorithms with and without point spread function (PSF) modeling. For both scanners, a post-reconstruction Gaussian filter of 3–7 mm in steps of 1 mm was applied. Attenuation correction was provided from a scaled computed tomography (CT) image of the phantom registered to the MR-based attenuation images and verified to align on the non-attenuation corrected PET images. For each of these image reconstruction parameter sets, contrast recovery coefficients (CRCs) were determined for the SUVmean, SUVmax and SUVpeak for each sphere. A hybrid metric combining the root-mean-squared discrepancy (RMSD) and the absolute CRC values was used to simultaneously optimize for best match in CRC between the two scanners while simultaneously weighting toward higher resolution reconstructions. The image reconstruction parameter set was identified as the best candidate reconstruction for each vendor for harmonized PET image reconstruction. Results The range of clinically relevant image reconstruction parameters demonstrated widely different quantitative performance across cameras. The best match of CRC curves was obtained at the lowest RMSD values with: for CRCmean, 2 iterations-7 mm filter on the GE Signa and 4 iterations-6 mm filter on the Siemens mMR, for CRCmax, 4 iterations-6 mm filter on the GE Signa, 4 iterations-5 mm filter on the Siemens mMR and for CRCpeak, 4 iterations-7 mm filter with PSF on the GE Signa and 4 iterations-7 mm filter on the Siemens mMR. Over all reconstructions, the RMSD between CRCs was 1.8%, 3.6% and 2.9% for CRC mean, max and peak, respectively. The solution of 2 iterations-3 mm on the GE Signa and 4 iterations-3 mm on Siemens mMR, both with PSF, led to simultaneous harmonization and with high CRC and low RMSD for CRC mean, max and peak with RMSD values of 2.8%, 5.8% and 3.2%, respectively. Conclusions For two commercially available PET/MRI scanners, user-selectable parameters that control iterative updates, image smoothing and PSF modeling provide a range of contrast recovery curves that allow harmonization in harmonization strategies of optimal match in CRC or high CRC values. This work demonstrates that nearly identical CRC curves can be obtained on different commercially available scanners by selecting appropriate image reconstruction parameters.

Quantitative Effect of Metal Artefact Reduction on CT-based attenuation correction in FDG PET/CT in patients with hip prosthesis

Background Metal artefact reduction (MAR) techniques still are in limited use in positron emission tomography/computed tomography (PET/CT). This study aimed to investigate the effect of Smart MAR on quantitative PET analysis in the vicinity of hip prostheses. Materials and methods Activities were measured on PET/CT images in 6 sources with tenfold activity concentration contrast to background, attached to the head, neck and the major trochanter of a human cadaveric femur, and in the same sources in similar locations after a hip prosthesis (titanium cup, ceramic head, chrome-cobalt stem) had been inserted into the femur. Measurements were compared between PET attenuation corrected using either conventional or MAR CT. In 38 patients harbouring 49 hip prostheses, standardized uptake values (SUV) in 6 periprosthetic regions and the bladder were compared between PET attenuation corrected with either conventional or MAR CT. Results Using conventional CT, measured activity decreased with 2 to 13% when the prosthesis was inserted. Use of MAR CT increased measured activity by up to 11% compared with conventional CT and reduced the relative difference with the reference values to under 5% in all sources. In all regions, to the exception of the prosthesis shaft, SUVmean increased significantly (p < 0.001) by use of MAR CT. Median (interquartile range) percentual increases of SUVmean were 1.4 (0.0–4.2), 4.0 (1.8–7.8), 7.8 (4.1–12.4), 1.5 (0.0–3.2), 1.4 (0.8–2.8) in acetabulum, lateral neck, medial neck, lateral diaphysis and medial diaphysis, respectively. Except for the shaft, the coefficient of variation did not increase significantly. Except for the erratic changes in the prosthesis shaft, decreases in SUVmean were rare and small. Bladder SUVmean increased by 0.9% in patients with unilateral prosthesis and by 4.1% in patients with bilateral prosthesis. Conclusions In a realistic hip prosthesis phantom, Smart MAR restores quantitative accuracy by recovering counts in underestimated sources. In patient studies, Smart MAR increases SUV in all areas surrounding the prosthesis, most markedly in the femoral neck region. This proves that underestimation of activity in the PET image is the most prevalent effect due to metal artefacts in the CT image in patients with hip prostheses. Smart MAR increases SUV in the urinary bladder, indicating effects at a distance from the prosthesis.